Jane Armstrong explains how a strict diet keeps her two-year-old son safe from the serious dangers of his rare liver condition
It is lunchtime. I am measuring out 110 grams of boiled potato (no skin), 30 grams of broccoli, and 48 grams of yoghurt for my second son, Nick. He will also be having as much low-protein topped apple crumble as he likes, and a few teaspoons of a very strong and unpleasant diet supplement. Unlike his elder brother, who is free to eat as many or as few of his fish fingers as he fancies, Nick has to eat up all of his measured food - no more and no less. Incredibly - to anyone familiar with two-year- olds - it's all, apart from the paste, going down without fuss.

Every child born in the UK is tested soon after birth for the condition which requires this diet, and yet it is so rare that relatively few people are aware of its existence. Although 1 in 60 people carries the gene defect involved, only 1 in 12,000 is born with the condition itself. The community midwife who administered Nick's heel-prick blood test cheerfully reassured us that no one ever came up positive.

Our healthy and extremely cheerful two-year-old has phenylketonuria (PKU to those who have to say it every day): his liver lacks sufficient quantities of an enzyme that breaks down phenylalanine, one of the major constituents of protein. When PKU is not treated with a reduced-protein diet, excess phenylalanine builds up in the brain and causes mental retardation, varying from mild to severe. If the condition is untreated in the first years of life, the brain damage can be irreversible.

The diet treatment for PKU was discovered only about 30 years ago; testing for the condition was introduced at about the same time. Before that, many sufferers ended up in mental institutions. The health visitor who discussed Nick's test result with us, when he was three weeks old, put this vividly (though in fact she was oversimplifying): when I asked her if Nick's father or I might have PKU without having realised it because we would never have been tested, she said: "If you did, you wouldn't be sitting here talking to me now."

Our baby, we were told, was one of the 50 or 60 children born with PKU every year; a strange contradiction awaited us. After the diagnosis, we were referred to the Children's Hospital at Great Ormond Street, London. We had to digest the shocking news that our child had a disease we'd never heard of, one that causes potentially severe mental retardation; but we were also told that on a strict diet he would be a perfectly normal child, and reach his full intellectual capacity.

Other people find this contradiction hard to grasp, too. Friends often ask whether Nick is "well". The answer is that he is not "ill" at all; but he is still different. What makes it even more difficult to pin down is that without a weekly blood test to check for levels of phenylananine in his blood, which we now carry out ourselves, there would be no indication that he had any problem for weeks or even months. There would be no "illness", just a gradual loss of concentration and abilities. In university students with the disease who leave off the diet for a few weeks, grades have been found to slip drastically; babies who are left untreated develop late, and a few may never learn to walk.

Food is such a central part of life that although PKU is invisible in terms of the child's health, it can never be invisible from day to day. The minute Nick sits down to a meal it is clear that he is not an ordinary child but a sort of super-vegan who doesn't even eat soya and lentils. We never go anywhere without our electronic scales. If Nick eats at someone else's house it means either preparing everything in advance and taking it with us or a detailed conversation with a bewildered parent about what he can eat. Having tea with a friend on impulse is very difficult. When we go on holiday it means taking an enormous box full of low-protein substitute biscuits and pasta, special vitamins and diet supplements. The supplements contain synthetic protein without phenylalaline, and added minerals.

We also have to make sure that we take our blood-testing kit so that we can do the test, post it off to be analysed for phenylalanine levels, and talk to the Great Ormond Street dietician when she rings to discuss Nick's diet for the week ahead. We also panic more about childhood illnesses: any fever or stomach upset sends the phenylalanine level shooting up, as the body starts to consume its own protein.

As a child with PKU grows older and the brain becomes more developed, the diet becomes less crucial, but it is still a diet for life, and the amount of protein allowed does not increase greatly with age. On a relatively mild diet, Nick is allowed the protein equivalent of eight Weetabix a day; many others of his age and much older are affected more severely.

Geneticists like PKU because it provides a textbook example of the way genetic mistakes work. Like other conditions - such as the more common, and more devastating, cystic fibrosis - it is carried on recessive genes, one from each parent.

They are beginning to trace different strains of PKU in different populations, put rough dates to their origin, and tie them into the history of population movements. If, for example, both sides of your family are Jewish and you have severe PKU, it is highly likely that you have a variant that started in one particular Yemeni family in the 14th century.

PKU is also likely to be a candidate for gene therapy, at least in the more severe cases. However, because the normal gene has to be introduced directly into the liver, the process is not straightforward.

The National Society for Phenylketonuria: 7 Southfield Close, Willen, Milton Keynes, MK15 9LL.